Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A mobile device for dynamically providing content to a user, comprising: a memory storing instructions; and a processor configured to execute the instructions to perform operations including: receiving a recommendation matrix associating potential offers with potential trigger events, the matrix based on a machine learning algorithm; receiving a travel graph specific to a user, the travel graph generated using location information provided by the mobile device and indicating travel movements of the mobile device, over a prior period of time, between designated geographical locations and a consumer location; receiving a trigger event associated with a geographical location determined from the travel graph; generating feature data based on the received trigger event, wherein the feature data indicates an update to a social network parameter of the user; identifying a first one of the offers using the feature data and the recommendation matrix; providing an indication of the first offer; receiving content associated with the first offer for display on the mobile device; and receiving redemption information of the first offer during a period of time, wherein the redemption information indicates success of the first offer, including at least one of: an amount of time between the display of the first offer and acceptance of the first offer; or a time elapsed up to a defined period.
A mobile device system dynamically provides personalized content to users based on their travel patterns and real-time contextual data. The system addresses the challenge of delivering relevant offers to users at optimal times by leveraging machine learning and location-based analytics. The device includes a processor and memory that execute operations to receive a recommendation matrix, which associates potential offers with trigger events using a machine learning algorithm. The system also receives a user-specific travel graph generated from historical location data, mapping movements between designated geographical locations and a consumer location. When a trigger event occurs at a specific location identified in the travel graph, the device generates feature data reflecting updates to the user's social network parameters. This data, combined with the recommendation matrix, identifies a relevant offer. The device then displays the offer and tracks redemption metrics, including the time between display and acceptance, or the elapsed time until a defined period expires. This approach ensures timely, contextually relevant offers, improving user engagement and offer effectiveness. The system dynamically adapts to user behavior, enhancing personalized marketing and service delivery.
2. The mobile device of claim 1 , wherein identifying the first offer includes calculating score values for the potential offers using the feature data and the recommendation matrix.
This invention relates to a mobile device system for personalized offer recommendations. The system addresses the challenge of efficiently selecting and presenting relevant offers to users based on their preferences and behavior. The mobile device includes a processor and memory storing instructions that, when executed, perform offer recommendation tasks. The device identifies a first offer from a set of potential offers by calculating score values for each offer using feature data and a recommendation matrix. The feature data may include user-specific information such as browsing history, purchase history, or demographic details, while the recommendation matrix represents relationships between offers and user features. The system then ranks the offers based on these scores and selects the highest-scoring offer for presentation. The device may also filter offers based on predefined criteria, such as offer expiration dates or user location, before scoring. The recommendation matrix is dynamically updated to reflect changes in user behavior or offer availability, ensuring recommendations remain relevant. This approach improves the accuracy and efficiency of offer recommendations, enhancing user engagement and satisfaction. The system may further include a display for presenting the selected offer and a communication interface for retrieving offer data from a remote server.
3. The mobile device of claim 2 , wherein calculating the score values includes: generating a feature vector based on the feature data, and multiplying the recommendation matrix with the feature vector to generate the score values.
This invention relates to a mobile device configured to provide personalized recommendations to a user. The device addresses the problem of efficiently generating relevant recommendations by leveraging user-specific feature data and a recommendation matrix. The mobile device includes a processor and memory storing instructions that, when executed, perform operations to calculate score values for items or content. These score values determine the relevance of recommendations for the user. The calculation involves generating a feature vector from the user's feature data, which may include preferences, behavior patterns, or contextual information. The feature vector is then multiplied by a recommendation matrix to produce the score values. The recommendation matrix is a precomputed data structure that encodes relationships between items and features, allowing for efficient scoring. This approach enables real-time or near-real-time recommendation generation while minimizing computational overhead. The invention improves recommendation accuracy and user experience by dynamically adapting to the user's evolving preferences and context. The mobile device may further refine recommendations by updating the feature vector or recommendation matrix based on user feedback or new data. This system is particularly useful in applications such as content streaming, e-commerce, or personalized advertising, where timely and relevant recommendations enhance user engagement.
4. The mobile device of claim 1 , wherein the recommendation matrix includes weight values for combinations of the potential offers and the potential trigger events.
A mobile device system provides personalized recommendations by analyzing user behavior and contextual data to suggest relevant offers or actions. The system identifies potential offers, such as promotions or notifications, and potential trigger events, such as user location or time of day, to determine optimal timing and content for recommendations. A recommendation matrix assigns weight values to combinations of these offers and trigger events, allowing the system to prioritize and select the most relevant suggestions based on historical user interactions and contextual factors. The matrix dynamically adjusts these weights to improve recommendation accuracy over time. The system may also filter recommendations based on user preferences, device capabilities, or external conditions, ensuring that only the most pertinent offers are presented. This approach enhances user engagement by delivering timely and personalized content, reducing irrelevant notifications, and improving overall user experience. The system may operate in real-time, continuously analyzing data to refine recommendations as the user's context changes.
5. The mobile device of claim 1 , wherein the operations further include sending the feature data to a remote computer.
A mobile device is configured to process sensor data to detect and analyze features of a physical environment. The device includes one or more sensors, such as cameras, microphones, or other input devices, that capture data from the environment. The device processes this sensor data to extract feature data, which may include visual, auditory, or other characteristics of objects, surfaces, or events in the environment. The extracted feature data is then transmitted to a remote computer for further analysis, storage, or use in applications such as augmented reality, navigation, or environmental monitoring. The device may also include a processor that executes instructions to perform these operations, ensuring efficient and accurate feature extraction. The remote computer may be part of a cloud-based system or a dedicated server, enabling scalable processing and access to the feature data. This system allows for real-time or batch processing of environmental data, improving situational awareness and enabling advanced applications that rely on detailed environmental analysis.
6. The mobile device of claim 5 , wherein the operations further include: receiving a second recommendation matrix from the remote computer, the second recommendation matrix updated using the feature data; and identifying a second offer using the feature data and the second recommendation matrix.
This invention relates to mobile devices that generate personalized offers based on user data and remote recommendations. The problem addressed is the need for dynamic, context-aware offer generation that adapts to user preferences and behavior in real time. The mobile device collects feature data, such as user interactions, device usage patterns, and environmental context, to generate a first recommendation matrix. This matrix is used to identify an initial offer tailored to the user. The device then transmits the feature data to a remote computer, which processes it to update the recommendation matrix. The updated matrix is sent back to the mobile device, which uses it along with the feature data to identify a second, more refined offer. This iterative process ensures offers remain relevant as user behavior and preferences evolve. The system leverages both local processing on the mobile device and cloud-based computation to balance responsiveness with computational efficiency. By continuously refining recommendations based on real-time data, the invention improves the accuracy and personalization of offers compared to static or infrequently updated systems. The approach is particularly useful in applications like targeted advertising, content recommendations, or personalized services where context-aware suggestions enhance user engagement.
7. The mobile device of claim 6 , wherein the operations further include: providing an indication of the second offer; and receiving second content associated with the second offer for display on the mobile device.
This invention relates to mobile devices configured to process and display offers, such as promotions or advertisements, to users. The problem addressed is the need for mobile devices to efficiently manage and present multiple offers, ensuring users receive relevant and timely information. The mobile device includes a processor and memory storing instructions that, when executed, perform operations to handle these offers. The device receives a first offer and associated first content, then displays the first content on the mobile device. The device also receives a second offer and associated second content, and provides an indication of the second offer while displaying the first content. Upon user interaction with the indication, the device replaces the first content with the second content for display. This allows users to seamlessly switch between offers without losing context. The invention improves user engagement by dynamically managing multiple offers and ensuring relevant content is prioritized. The system ensures smooth transitions between offers, enhancing the user experience by reducing interruptions and maintaining focus on the most relevant information.
8. The mobile device of claim 1 , wherein the feature data further indicates at least one of: a geographic location of the mobile device; that the mobile device is traveling; that the mobile device was used to generate a search query; or that the mobile device is located within a determined distance from a specified geographic location.
A mobile device system collects and processes feature data to enhance user interactions. The system gathers information about the device's state and context, including geographic location, movement status, search query generation, and proximity to specific locations. By analyzing this data, the device can provide more relevant services, such as location-based recommendations, personalized search results, or context-aware notifications. The system dynamically adjusts its functionality based on real-time conditions, improving user experience by tailoring responses to the device's current situation. For example, if the device detects movement, it may prioritize navigation-related features, while proximity to a predefined location could trigger location-specific content or alerts. This approach ensures that the device adapts to the user's needs without requiring manual input, streamlining interactions and reducing cognitive load. The system leverages sensor data, network signals, and user activity logs to derive meaningful insights, enabling proactive and contextually appropriate responses. This enhances efficiency and usability, particularly in scenarios where immediate, relevant information is critical.
9. The mobile device of claim 1 , wherein the operations further include generating at least one feature data event based on the travel graph.
A mobile device system is designed to analyze and optimize travel routes by generating a travel graph representing movement patterns. The system collects sensor data from the mobile device, such as GPS, accelerometer, and gyroscope readings, to construct a travel graph that maps the device's movement over time. This graph includes nodes representing locations and edges representing transitions between those locations, along with associated metadata like timestamps, speed, and direction. The system processes this travel graph to identify travel patterns, such as frequent routes or common stops, and uses this information to improve navigation, predict travel times, or suggest optimized routes. Additionally, the system generates feature data events based on the travel graph, which can include derived metrics like travel efficiency, route deviations, or activity recognition (e.g., walking, driving, or public transit). These events can be used for further analysis, such as personalizing travel recommendations or detecting anomalies in movement behavior. The system may also integrate with external data sources, such as traffic or weather information, to enhance the accuracy of travel predictions. The overall goal is to provide a more intelligent and adaptive travel experience by leveraging real-time and historical movement data.
10. The mobile device of claim 1 , wherein the content includes content associated with at least one of an advertisement, a promotional offer, a discount, a coupon, and a sale associated with an item or service.
This invention relates to mobile devices configured to display content related to advertisements, promotional offers, discounts, coupons, or sales for items or services. The mobile device includes a display screen for presenting such content to a user. The content may be dynamically generated or retrieved based on user preferences, location, or other contextual factors. The device may also include processing capabilities to analyze user interactions with the displayed content, such as tracking clicks, purchases, or engagement metrics. Additionally, the mobile device may support features like barcode scanning, QR code reading, or direct purchase options to facilitate transactions tied to the promotional content. The system may integrate with external databases or servers to fetch real-time offers or update pricing information. The goal is to enhance user engagement with commercial content while providing seamless access to deals and promotions.
11. A computer-implemented method performed by a mobile device for dynamically providing content to a user, comprising: storing a recommendation matrix associating potential offers with potential trigger events, the matrix based on a machine learning algorithm; receiving a travel graph specific to a user, the travel graph generated using location information provided by the mobile device and indicating travel movements of the mobile device, over a prior period of time, between designated geographical locations and a consumer location; identifying a trigger event associated with a geographical location, the geographical location determined from the travel graph; generating feature data based on the received trigger event, wherein the feature data indicates an update to a social network parameter of the user; identifying a first one of the offers using the feature data and the recommendation matrix; providing an identification of the first offer; receiving content associated with the first offer for display on the mobile device; and receiving redemption information of the first offer during a period of time, wherein the redemption information indicates success of the first offer, including at least one of: an amount of time between the display of the first offer and acceptance of the first offer; or a time elapsed up to a defined period.
This invention relates to a mobile device system for dynamically providing personalized content, such as offers, to users based on their travel patterns and real-time contextual data. The system addresses the challenge of delivering relevant offers to users at optimal times by leveraging machine learning and location-based tracking. The method involves storing a recommendation matrix that links potential offers to specific trigger events, where the matrix is generated using a machine learning algorithm. The system receives a user-specific travel graph, which is constructed from location data collected by the mobile device over a prior period. This travel graph maps the user's movements between designated geographical locations and their consumer location, allowing the system to identify relevant trigger events based on the user's current or predicted location. When a trigger event is detected, the system generates feature data, which includes updates to the user's social network parameters. This feature data is then used in conjunction with the recommendation matrix to identify the most suitable offer for the user. The system provides the offer to the user for display on their mobile device and tracks redemption information, including metrics such as the time between offer display and acceptance, or the elapsed time until a defined period expires. This data helps refine future recommendations by assessing the effectiveness of the offer. The system ensures timely and contextually relevant offers, enhancing user engagement and conversion rates.
12. The method of claim 11 , wherein the identification of the first offer is provided to an advertisement server and the content associated with the first offer is received from the advertisement server.
The invention relates to digital advertising systems, specifically methods for identifying and delivering targeted offers to users. The problem addressed is the inefficiency in dynamically selecting and presenting relevant advertisements to users based on their behavior or preferences. The method involves identifying a first offer from a set of available offers, where the selection is based on predefined criteria such as user behavior, location, or historical data. The identification of this first offer is then communicated to an advertisement server, which processes the request and retrieves the corresponding content associated with the offer. The advertisement server returns this content, which may include multimedia elements like images, videos, or interactive components, to be displayed to the user. This approach ensures that the most relevant and timely offers are dynamically delivered to users, improving engagement and conversion rates. The method may also involve additional steps such as tracking user interactions with the displayed content, updating user profiles based on these interactions, and refining future offer selections accordingly. The system may further include mechanisms to prioritize offers based on real-time data, ensuring that the most impactful advertisements are presented to users. By leveraging an advertisement server, the system efficiently manages and distributes content, optimizing the delivery process for both advertisers and users.
13. The method of claim 11 , the method further comprising decrypting an encrypted recommendation matrix to generate the recommendation matrix.
A system and method for generating personalized recommendations using a recommendation matrix. The technology addresses the challenge of efficiently providing tailored suggestions to users based on their preferences or behavior, particularly in environments where data privacy or security is a concern. The method involves processing user data to create a recommendation matrix, which encodes relationships between users and items or content. This matrix is then used to generate recommendations by identifying items that are likely to be of interest to a specific user. To enhance security, the recommendation matrix may be encrypted to protect sensitive information. The method includes decrypting the encrypted recommendation matrix to restore its original form, enabling the generation of accurate recommendations while maintaining data confidentiality. The decryption process ensures that the recommendation system can operate on the unencrypted matrix without exposing the underlying data to unauthorized access. This approach balances the need for personalized recommendations with the requirement to safeguard user privacy and data integrity. The system may be applied in various domains, including e-commerce, content streaming, and social media, where personalized recommendations are valuable for improving user engagement and satisfaction.
14. The method of claim 11 , wherein the recommendation matrix comprises rows correlating to potential offers and columns correlating to potential trigger event data.
This invention relates to systems for generating personalized recommendations based on trigger events. The problem addressed is the need for dynamic, context-aware recommendations that adapt to real-time user behavior or external conditions. The invention provides a method for creating a recommendation matrix that organizes potential offers and trigger event data to improve recommendation accuracy. The recommendation matrix is structured with rows representing potential offers and columns representing potential trigger event data. Each cell in the matrix contains a value indicating the relevance or likelihood of a specific offer being recommended when a corresponding trigger event occurs. Trigger events may include user actions, time-based conditions, or external factors like weather or location. The matrix is dynamically updated based on user interactions, historical data, or real-time analytics to ensure recommendations remain relevant. The method involves analyzing trigger event data to identify patterns or correlations with user preferences. Machine learning techniques may be used to refine the matrix over time, improving the accuracy of recommendations. The system can then generate personalized offers by querying the matrix when a trigger event is detected, selecting the most relevant offers based on the stored correlations. This approach enhances recommendation systems by making them more responsive to contextual factors, leading to higher user engagement and satisfaction. The matrix structure allows for efficient computation and scalability, making it suitable for large-scale applications.
15. The method of claim 14 , wherein the recommendation matrix comprises model coefficients associated with a potential trigger event and a potential offer.
A system and method for generating personalized recommendations in a digital platform, particularly for triggering offers or actions based on user behavior. The technology addresses the challenge of delivering relevant and timely recommendations to users in dynamic environments, such as e-commerce, advertising, or content delivery, where user preferences and contextual factors influence engagement. The method involves analyzing user interactions and contextual data to identify patterns and preferences, which are then used to generate a recommendation matrix. This matrix includes model coefficients that quantify the relationship between potential trigger events (e.g., user actions, time of day, device type) and potential offers (e.g., discounts, content suggestions, or service promotions). The coefficients represent the likelihood or desirability of a user responding positively to an offer when a specific trigger event occurs. The system dynamically updates the recommendation matrix based on real-time data, ensuring recommendations remain relevant. The method may also incorporate machine learning techniques to refine the coefficients over time, improving recommendation accuracy. By leveraging this matrix, the system can predict the most effective offers to present to users, enhancing user experience and engagement. The approach is particularly useful in environments where personalized, context-aware recommendations are critical for driving user actions.
16. The method of claim 11 , wherein the feature data comprises a feature vector, and generating the feature data based on the received trigger event comprises applying current event data to rules, the rules corresponding to components of the feature vector.
This invention relates to a system for processing trigger events in a technical domain, particularly for generating feature data used in decision-making or analysis. The problem addressed is the need to efficiently and accurately transform raw event data into structured feature data that can be used for further processing, such as machine learning or rule-based decision systems. The method involves receiving a trigger event, which is an input signal or data indicating an occurrence that requires processing. The trigger event may originate from various sources, such as sensors, user inputs, or system logs. The method then generates feature data based on the trigger event, where the feature data is structured as a feature vector—a numerical representation of the event's characteristics. This feature vector is composed of multiple components, each corresponding to a specific aspect of the event. To generate the feature vector, the method applies the current event data to a set of predefined rules. These rules define how the raw event data should be transformed into the components of the feature vector. For example, a rule might specify that a particular sensor reading should be normalized and assigned to a specific position in the feature vector. The rules ensure consistency and relevance in the feature data, making it suitable for downstream applications like predictive modeling or real-time decision-making. The system dynamically processes trigger events, ensuring that the feature data remains up-to-date and reflective of the current state of the system. This approach improves the accuracy and efficiency of subsequent analyses or actions based on the feature data.
17. The method of claim 16 , wherein identifying the first offer comprises: multiplying the recommendation matrix by the feature vector to generate a score matrix; determining a highest-score value for the potential offers in the score matrix; selecting an offer of the potential offers corresponding to the highest-score value for the potential offers in the score matrix as the first offer.
This invention relates to a method for identifying and selecting offers, such as recommendations or promotions, based on a recommendation matrix and feature vectors. The method addresses the challenge of efficiently determining the most relevant offer from a set of potential offers by leveraging matrix operations and scoring mechanisms. The process involves generating a score matrix by multiplying a recommendation matrix with a feature vector. The recommendation matrix represents relationships between users and potential offers, while the feature vector encodes characteristics of a user or context. The score matrix contains values indicating the suitability of each potential offer for the user. The highest-score value in the score matrix is identified, and the corresponding offer is selected as the most relevant recommendation. This approach ensures that the selection process is computationally efficient and scalable, as it relies on matrix multiplication and simple value comparison rather than exhaustive evaluation of each offer. The method is particularly useful in systems where personalized recommendations or dynamic offers are generated in real-time, such as e-commerce, advertising, or content delivery platforms. By optimizing the selection process, the invention improves the accuracy and speed of offer recommendations, enhancing user experience and engagement.
18. The method of claim 16 , wherein the components of the feature vector are binary-valued.
This invention relates to a method for processing data using feature vectors, specifically where the components of the feature vector are binary-valued. Feature vectors are used in various applications, such as machine learning, pattern recognition, and data compression, to represent data in a structured format. A common challenge in these applications is efficiently encoding and processing feature vectors while maintaining computational efficiency and accuracy. The method involves generating a feature vector from input data, where each component of the vector is a binary value (either 0 or 1). Binary-valued feature vectors simplify computations, reduce memory usage, and improve processing speed compared to vectors with continuous or multi-valued components. The binary nature of the components allows for efficient storage and fast operations like bitwise logic, which are particularly useful in hardware implementations or real-time systems. The method may also include steps for extracting features from the input data, normalizing the features, and converting them into binary values. Techniques such as thresholding, hashing, or quantization can be used to convert non-binary features into binary form. The resulting binary feature vector can then be used in further processing, such as classification, clustering, or dimensionality reduction. By using binary-valued feature vectors, the method enables faster and more resource-efficient data processing while maintaining the necessary accuracy for applications like image recognition, natural language processing, and anomaly detection. The approach is particularly beneficial in environments with limited computational resources or where real-time performance is critical.
19. The method of claim 11 , wherein the feature data indicates at least one of: a geographic location of the mobile device; that the mobile device is traveling; that the mobile device was used to generate a search query; or that the mobile device is located within a determined distance from a specified geographic location.
This invention relates to mobile device tracking and data analysis, specifically for determining and utilizing contextual information about a mobile device's status and behavior. The method involves collecting feature data from a mobile device to assess its current state, such as geographic location, movement status, search query generation, or proximity to a predefined location. This data is used to infer contextual insights, such as whether the device is stationary, in transit, or engaged in specific activities like searching for information. The system processes this feature data to generate actionable intelligence, which can be applied in various applications, including targeted advertising, location-based services, or user behavior analysis. The method ensures that the feature data is dynamically updated to reflect real-time changes in the device's status, enabling accurate and timely decision-making. By analyzing these contextual cues, the system can provide personalized or contextually relevant responses, improving user experience and service efficiency. The invention addresses the challenge of deriving meaningful insights from mobile device data to enhance service delivery and user engagement.
20. A system for recommendation data associated with offers to be presented to a user, comprising: a memory storing instructions; and a processor configured to execute the instructions to perform operations including: receiving a travel graph specific to a user, the travel graph generated using location information provided by the mobile device and indicating travel movements of the mobile device, over a prior period of time, between designated geographical locations and a consumer location; generating a recommendation matrix based on the feature data provided by the mobile device, consumer data for the user provided by a third party, offer data for potential offers, and a machine learning algorithm, the recommendation matrix including model coefficient values associating trigger events and the potential offers, at least one trigger event based on the travel graph, wherein the feature data indicates an update to a social network parameter of the user; sending the recommendation matrix to the mobile device for use in generating a score matrix for identifying a first one of the offers of the potential offers to display on the mobile device; and receiving redemption information of the first offer during a period of time, wherein the redemption information indicates success of the first offer, including at least one of: an amount of time between the display of the first offer and acceptance of the first offer; or a time elapsed up to a defined period.
This system provides personalized offer recommendations to users based on their travel patterns and other data. The system addresses the challenge of delivering relevant offers to users by analyzing their movement between locations and other contextual information. A travel graph is generated from location data collected by a user's mobile device, tracking movements between designated geographical locations and a consumer location over a prior period. The system also gathers feature data from the mobile device, including updates to social network parameters, and consumer data from third-party sources. Offer data for potential offers is combined with this information using a machine learning algorithm to generate a recommendation matrix. This matrix contains model coefficient values that associate trigger events, derived from the travel graph, with potential offers. The recommendation matrix is sent to the mobile device, which uses it to generate a score matrix for identifying the most relevant offer to display. The system tracks redemption information for the displayed offer, including metrics such as the time between display and acceptance, to measure the offer's success. This approach enhances the relevance of offers by leveraging real-time and historical user data, improving user engagement and offer conversion rates.
Unknown
September 3, 2019
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.